INDEX

Explanations

terms related to authority and compliance

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

usan

-0.17

åįĵ

-0.15

leted

-0.15

ê²°

-0.14

UE

-0.14

à¸°

-0.14

alink

-0.14

à¤°à¤²

-0.14

adil

-0.13

æģ¯

-0.13

POSITIVE LOGITS

ÑĢÐ¾ÑĦ

0.15

gan

0.15

assic

0.15

hyp

0.15

essen

0.14

æķ·

0.14

Ãªu

0.14

.eclipse

0.14

AMPLE

0.14

oga

0.13

Activations Density 0.004%

terms related to authority and compliance

No Comments

No Known Activations